Bid answers are scored on specific, evidenced claims about past projects and people. A tool that writes fluent generalities loses marks; one that finds the right project and quotes it accurately is worth having.
AI bid writing for architecture and engineering firms means drafting tender and framework quality answers from the firm’s own past submissions, project sheets and CVs, mapped to the evaluation criteria and word limits, so the bid team edits and strengthens a draft instead of searching for material.
Public work in the UK now runs under the Procurement Act 2023, in the EU under Directive 2014/24/EU and its national laws (in Spain, Ley 9/2017 de Contratos del Sector Público). The formats differ; the burden is the same. Each tender asks for methodology, relevant experience, team, social value and quality management, scored against published criteria, often in a few days. The material exists in last year’s bids and in project sheets nobody has updated.
The risk is specific. A model asked to write about relevant experience will happily invent a project, a value or a completion date. In a public tender that is a misrepresentation, so the system only writes claims it can trace to a document in your archive, and shows the source beside each one.
We rerun past tenders where you hold the feedback and scores. A draft is judged on whether it covers the points evaluators rewarded, and whether every factual claim in it traces to a source. The second is checked automatically on every draft, not sampled, because a single invented reference project can disqualify a bid.
The archive needs some care first: project sheets with current values and dates, CVs with permission to reuse, and a note of which past answers scored well. That work pays off beyond bids. Submission through the buyer’s portal, and the declarations that go with it, remain with the person authorised to make them. Contractors bidding for construction packages have a related problem, covered under bids and estimating.
ITT, criteria, forms
questions · weights · limits
bids · project sheets · CVs
every claim sourced
and the authorised person submits
shown as a gap, never filled
It can draft one, and the quality depends almost entirely on the evidence it draws from. With a well-kept archive of past bids, project sheets and CVs, it produces specific, criteria-mapped drafts that a bid team can improve quickly. Without that archive it produces fluent generalities that score poorly. Either way, people edit, approve and submit.
Generally yes, but check each tender. Some buyers now ask suppliers to declare AI use or restrict it, and every buyer expects the content to be accurate and the declarations true. The supplier is responsible for what is submitted regardless of how it was drafted, which is why our system will not write claims it cannot trace to a document.
By only writing from retrieved documents and checking every factual claim against them before the draft is shown. Project names, values, dates and people must match a source in your archive or the sentence is removed and the gap flagged. That check runs on every draft, because a single invented reference can disqualify a bid.
What we built, what stage it is at, and a link to the live thing wherever there is one.
yourOS is an operating system with an AI at its heart. You talk to it like a person, out loud or typed. It makes your calls, plans your days and works on your projects with a crew of agents, on your own private server. Our own product, built from the first commit; the waitlist is open.
Website, backend and app assets for Sonora, a free AI sound-therapy app on iOS and Android published by Cognitive AI.
Brand, website, prototype, apps on five platforms, the AI backend and the algorithm-intelligence pipeline for Podswap, a social growth engine that runs without a human in the loop.
In closed beta: Roxy, a private AI assistant with permanent memory running on infrastructure the user owns. Brand, site, product, backend and conversational design.
Live at coherence.com: brand identity generated with AI agents we directed, the motion system and website for the group behind Sonora, AboutAI, WakeUp and MudShadows. Agent work remains in build.
Live at domainbooth.com for DomainBooth, a premium domain brokerage: a new animated mark, every page redesigned and rebuilt with designed mobile layouts, and a custom CMS for its sales record, featured names, press and enquiries.
Bring a past tender with its feedback; we will draft it again from your archive and show where the evidence was thin.
30-minute technical call · no deck